In Layered Architecture as an example of a distributed environment, a Media Gateway Controller (MGC) typically uses a Gateway Control Protocol (GCP, H.248, Megaco) to control the set-up and release of bearers and the use of resources within the Media Gateway (MGW) to handle the user plane. Resources within the Media gateway may comprise echo cancellers, tone senders, tone receivers, transcoders, announcement machines, etc.
For the purpose outlined above, the MGC may use a set of call related commands, such as ADD, MOV, MOD, SUB to create, move, modify or remove a termination or a Notify command to inform the Media Gateway Controller MGC of an event detected by the Media Gateway MGW, e.g. a bearer has been established towards a peer MGW.
Furthermore, the MGC may use non-call related commands. These commands comprise a Service_Change, Audit_Value or Audit_Capabilities command.
Hence, in Layered Architecture the MGW is context and termination oriented whereas the MGC is call oriented.
In general terms, the Media Gateway Controller is part of the call control plane typically handling the call control protocols such as BICC, ISUP, RANAP, BSSAP, . . . whereas the Media Gateway is part of the user plane typically handling bearer control protocols such as Nb framing, RTP, IP, Q.AAL2, AAL2, . . . .
Although such a layered architecture provides numerous advantages, it is also a consequence that it becomes more difficult to trace calls and/or to debug problems once they occur in the layered architecture than it is in a traditional non-layered architecture.
When a problem occurs in layered architecture typically only a partial view of the history of events that might have lead to the problem is available to operators, maintenance personnel or trouble shooters.
The following two examples illustrate two exemplary scenarios.
Scenario I, a Fault Occurs within the Call Control Plane.
Assume a problem have been detected within a Media Gateway Controller (MGC). This problem typically leads to a release of the call by the Media Gateway Controller. At that stage the Media Gateway Controller will typically perform two actions:                Clear the call, i.e. all resources used by the call will be released, so that they become available for other calls. Such a clearing includes also resources used in the Media Gateway, i.e. SUB commands will be sent to all terminations involved in the Media Gateway.        Dump data available within the Media Gateway Controller so that it can be used for debugging purposes.        
However, the Media Gateway not having the call control plane view is not able to differentiate these SUB commands issued to an abnormal release from a SUB corresponding to a normal call release.
Hence, the Media Gateway is not aware that a problem related to this context and termination has occurred and consequently the MGW will typically erase any information that it might have kept during the existence of the context.
As a consequence, the Media Gateway will not be able to provide any information (logs, dumps) that would help to debug a problem that occurred or was detected in the Media Gateway Controller.
Scenario II, a Fault Occurs within the User Plane
Assume a problem have been detected within a Media Gateway. A consequence might be that the Media Gateway will not be able to execute certain subsequent commands. For example, the Media Gateway receives a MOD command. If the MOD command could not be executed an error message is replied to the Media Gateway Controller. Subsequently, the Media Gateway Controller typically releases the call.
Since, the call release is effected by an external reason, the MGC may consider that it doesn't need to dump any information that it might have kept during the existence of the call, but simply proceed with a normal release sequence.
As a consequence, the Media Gateway Controller will not be able to provide any information (logs, dumps) that would help to debug a problem that occurred or was detected in the MGW.
Furthermore, other contexts in the same or in another MGW as it might happen for example within a conference call may be also related to the call. Those contexts will naturally be removed when the call is cleared.
Hence, also the same or another Media Gateway related to the call will not be able to provide any information (logs, dumps) that would help to debug a problem that occurred or was detected in the Media Gateway.